Out of the tropics, but how? Fossils, bridge species, and thermal ranges in the dynamics of the marine latitudinal diversity gradient

Abstract

Latitudinal diversity gradients are underlain by complex combinations of origination, extinction, and shifts in geographic distribution and therefore are best analyzed by integrating paleontological and neontological data. The fossil record of marine bivalves shows, in three successive late Cenozoic time slices, that most clades (operationally here, genera) tend to originate in the tropics and then expand out of the tropics (OTT) to higher latitudes while retaining their tropical presence. This OTT pattern is robust both to assumptions on the preservation potential of taxa and to taxonomic revisions of extant and fossil species. Range expansion of clades may occur via "bridge species," which violate climate-niche conservatism to bridge the tropical-temperate boundary in most OTT genera. Substantial time lags (∼5 Myr) between the origins of tropical clades and their entry into the temperate zone suggest that OTT events are rare on a per-clade basis. Clades with higher diversification rates within the tropics are the most likely to expand OTT and the most likely to produce multiple bridge species, suggesting that high speciation rates promote the OTT dynamic. Although expansion of thermal tolerances is key to the OTT dynamic, most latitudinally widespread species instead achieve their broad ranges by tracking widespread, spatially-uniform temperatures within the tropics (yielding, via the nonlinear relation between temperature and latitude, a pattern opposite to Rapoport's rule). This decoupling of range size and temperature tolerance may also explain the differing roles of species and clade ranges in buffering species from background and mass extinctions.

Keywords: biodiversity; biogeography; climate; macroecology; macroevolution.

Fig. 1.

Species richness of present-day marine bivalves occurring at continental shelf depths (< 200 m).

Fig. 2.

First occurrences of bivalve genera in the tropical and extratropical fossil record for three successive geologic time bins. All pairs differ significantly (P = 0.001, exact binomial tests).

Fig. 3.

Spatial and temporal dynamics underyling the OTT process. (A) Lag between time of tropical origination (geologic age of genus) and entry into the temperate-zone fossil record (lag =distance below 1:1 diagonal) in North America, using only taxa that today extend at least as far as North Carolina or California (n = 22, Middle Miocene-Pleistocene genera). (B) For genera today occurring in both tropics and extratropics, the number of bridge species equals or exceeds the number of entirely extratropical species, regardless of genus age for 5-My cohorts back to 60–65 My. Diagonal = 1:1 line. (C) Minimum (white) and maximum (shaded) SSTs occupied today by tropics-only and extratropics-only species in the West Pacific compared with the tropical and extratropical parts of bridge species ranges. See Fig. S5 for further examples. (D) Percentage of genera today endemic to the New World tropics but having extratropical fossil occurrences, as a function of the geologic ages of the genera.

Fig. 4.

(A) Geographic ranges of bivalve species (white boxes) encompass narrower temperature ranges than expected from a null model (gray boxes; see SI for details); IQR, interquartile range. (B) Mean annual SSTs encompassed by geographic ranges-size classes of bivalve species, showing that widespread species tend to occupy warm temperatures.